Restaking takes ETH that is already staked on Ethereum and commits it to secure other protocols as well, adding a second layer of rewards on top of base staking income. The trade-off is plain: more yield comes with more ways to lose capital. That mix has made restaking one of the biggest themes in crypto and one of the least well understood.
In the source material, ordinary staking is the starting point. ETH is locked to help validators secure Ethereum, earning roughly 3% to 4% a year. If validators act dishonestly or fail to meet protocol rules, part of that stake can be slashed. Restaking extends the same economic security to other services, such as data availability systems, bridges, and oracles, so those protocols do not need to build their own token-based security model from scratch.
EigenLayer turned shared security into a major DeFi system
The concept was introduced by EigenLayer in 2023. By 2026, according to the source, it had grown into one of the largest systems in decentralized finance, with tens of billions of dollars committed. The appeal is straightforward: new protocols can tap into Ethereum-backed economic security on day one instead of spending time and capital trying to bootstrap trust on their own.
Three roles define how the model works. Restakers provide ETH or staked-ETH assets. Operators run the software required by the services being secured. The services themselves are called AVSs, short for Actively Validated Services. Most users do not run infrastructure directly; they delegate to operators, who usually keep around 10% of the rewards earned on behalf of restakers.
Why liquid restaking became the default path for users
Direct restaking is operationally heavy, which is why liquid restaking has become the main entry point. Users deposit into a protocol that manages operator selection, AVS exposure, and on-chain execution, then receive a liquid restaking token, or LRT, as a claim on that position. The source highlights leading examples including ether.fi’s eETH, Renzo’s ezETH, Kelp DAO’s rsETH, and Puffer’s pufETH.
The value of the liquid model is not hard to see. A standard restaking position is locked, but an LRT can be traded and reused across DeFi while still accumulating layered rewards. That composability pushed restaking far beyond a niche yield strategy. It also moved more risk into token wrappers and interconnected protocols, which means tradability improved without removing the underlying exposure.
How the yield stack forms, and how leverage changes the profile
The source gives a simple example with 10 ETH deposited into a liquid restaking protocol such as ether.fi. The first layer of return comes from Ethereum staking at around 3% to 4%. The second comes from AVS rewards, typically another 1% to 2%. A third layer, common in 2026, comes from points programs or token incentives. Put the real yield layers together and the result lands around 4% to 7%.
Some users push the structure much harder by looping. They post an LRT as collateral on a lending protocol, borrow more ETH, buy more LRT, and repeat. The source says headline returns can rise to 12% to 20% this way. The downside is equally clear: if the LRT trades below the value of the ETH it represents, or if market moves trigger liquidations, losses can expand quickly. This is leveraged yield farming, not passive income dressed in cleaner packaging.
Slashing, contract risk, and concentration remain central concerns
Restaking adds risk because it adds obligations. Slashing is the most direct one: users accept extra penalty conditions when they opt in to secure more services, and an operator’s poor execution can damage delegated capital. Smart-contract risk sits beside it. Funds can pass through Ethereum staking contracts, EigenLayer contracts, and the contracts of a liquid restaking protocol, so each added layer creates another point of failure.
Liquidity pressure matters too. In stressed markets, an LRT can trade below the value of its backing ETH, forcing exits at a discount. The source also points to concentration risk as billions of dollars flow into the sector: a small number of protocols and tokens control a large share of restaked ETH. If one heavily used component fails, the effect may not stay isolated for long.
At its core, restaking does not create yield out of nothing. It reuses the same capital across multiple security commitments, which means rewards stack only because risk stacks as well.

